AEGIS open-sources an MCP gateway requiring per-call approval of exact tool arguments
Leather_Tour_794 · reddit · 2026-10-27
The maintainer of AEGIS (MIT-licensed, Windows-first desktop beta from Escape Studio) argues that approving agent tool calls by name alone is insufficient—the same file-writing tool differs greatly between a notes path and a credential path. Its validating MCP gateway verifies the tool and schema on configured routes, then requires a one-use approval bound to the exact call and arguments; the grant is consumed before forwarding and separate requests need separate approvals. It also ships a selected-action policy that can deny routed commands pre-process-creation, static review of skills/projects, process investigation, and manual pause/stop. Controls require explicit setup; ordinary monitoring never auto-blocks. Code is on GitHub, and the author asks what additional context reviewers would want when approving calls.
Related event: Escape Studio Open-Sources AEGIS, an MCP Security Gateway(2 posts)→
More from coding & agent
- FrontierCS 2 launches on Kaggle with DeepMind: 48 tasks ask agents to beat published papers — lmoroney · 2026-10-27
- SkillHarm benchmark shows agent skill-poisoning attacks succeed up to 86.3% of the time — hhsun1 · 2026-10-27
- Dev builds Dreamcast game in 3 days with OpenAI's Astra, chasing locked 60fps — AIandDesign · 2026-10-27
- Dev builds Unreal prototype entirely via Claude Code CLI, never opening the editor — majidmanzarpour · 2026-10-27
- kimo: an open-source headless video editor driven by AI agents over MCP — LieTop7965 · 2026-10-27
- Dev breaks down $253 bill: 56 scaling sessions, one full test run only cost $4.50 — gdechichi · 2026-10-27